Hi,
I haven't had a chance to dig through
why this is coming up, but it seems
like count() for selectables is
getting rebound from what FromClause
defines. Test code and output are
below.
--- test code:
# Test count() on selectable
import sqlalchemy as rdb
engine = rdb.create_engine('sqlite:/
On Feb 2, 2006, at 3:04 PM, Michael Bayer wrote:
the mapper() object allows you to specify a primary_key on the mapper,
as it relates to its selectable:
s = mytable.select(table.c.col1==5)
m = mapper(s, primary_key=[s.c.id])
so you can stick a straight table or an aliased sele
On Feb 2, 2006, at 2:13 PM, Michael Bayer wrote:
you want to set primary keys on a selectable ? selectables are
supposed to very closely relate to SQL constructs. the mapper level
is where you can define your not-quite-SQL constructs and how they map
to your SQL constructs. since a SELECT st
hi,
i'm trying to use an aliased selectable
with a mapper, and the i'm getting an
error that the selectable does not have
a primary key.
the error also says that i can specify
the primary key to the mapper init, but
i'd rather tell the selectable what its
primary key is instead.
i see that ther
pretty sure the following was what was intended:
Index: sql.py
===
--- sql.py (revision 879)
+++ sql.py (working copy)
@@ -465,7 +465,7 @@
if obj is None:
if operator != '=':
On Jan 28, 2006, at 2:52 PM, Michael Bayer wrote:
What this is doing is taking the usual dependency sort at the table
level, detecting if there are cycles, and if so breaking it down to a
dependency sort at the object instance level. So you still need to
ultimately not have a circular relation
On Jan 26, 2006, at 8:47 AM, Michael Bayer wrote:
I like that patch, but just curious, how does this save memory ? If
the
object still exists in the identity map, which is a
WeakValueDictionary,
that implies that a reference is still hanging around on that object
somewhere else. Removing fro
hi,
i'm having some issues with memory
even with the weakref stuff. i have
a feeling there must be better solutions,
but for now i've given myself a manual
way of clearing out the identity_map.
i think i have the checks to make sure
that it's okay to clear the id correct,
but i'm not so sure. in
On Jan 24, 2006, at 4:29 PM, Michael Bayer wrote:
but show me what youre trying to do again, a little more simply, so i
can see what youre looking for.
hi mike,
thanks as always for your response.
i unfortunately haven't had a chance
to isolate some of the issues that
i'm currently having into
On Jan 24, 2006, at 1:51 PM, Michael Bayer wrote:
you should be able to do this:
del objectstore.uow().identity_map[objectstore.instance_key(obj)]
hi mike,
i couldn't seem to get this to work.
afaict the identity_map is empty with
the test code that i have - i looked
around just in case in uo
On Jan 23, 2006, at 8:56 PM, Michael Bayer wrote:
do this:
m = mapper(cls, table, is_primary=True)
thanks mike.
is there a better way to do
what i'm doing in the REGET
section?
---
if REGET:
del x1
x1 = m1.selectone()
---
ideally i'd like something like:
m1.clear_cache(x1)
o
hi,
i'm trying to do something which is
perhaps fundamentally ill-advised,
but it would be great if something
like this could work.
the code is below, but it probably
could use some explanation.
i start with 2 tables 't1' and 't2a',
where 't2a' FKs into 't1'. naturally,
i create a relation betw
On Jan 23, 2006, at 5:41 PM, Michael Bayer wrote:
hmm does it really ? I am most accustomed to Oracle where getting
just the first row back of a huge result set is not much different
than just getting one (and also, Oracle doesnt even have 'limit', and
I have instead hacked together some "rowi
On Jan 23, 2006, at 2:50 PM, [EMAIL PROTECTED] wrote:
also curious why you have a separate step for
generating primary keys, since that behavior is built-in via
sequences or
"defaults".
hmm; i was having some issues with that
before, but you're right, it seems to
work without. i'm not sure wh
the following minor patch makes
selectone() much faster for large
result sets:
---
Index: mapping/mapper.py
===
--- mapping/mapper.py (revision 839)
+++ mapping/mapper.py (working copy)
@@ -416,6 +416,7 @@
def selectone(self
On Jan 23, 2006, at 1:55 PM, Michael Bayer wrote:
So they are adding some
overhead to the commit, but not a lot. They are definitely not being
manifest as any additional SQL or anything like that.
hi mike,
thanks as always for the response.
i do understand (and see) that the actual sql
sent t
On Jan 20, 2006, at 9:51 PM, Michael Bayer wrote:
update and try it again. the fix i made for you yesterday was
slightly incorrect. youll notice that the dependency graph produced
for the second commit is much smaller after you update this change.
hi mike,
thanks for the update -
things se
On Jan 20, 2006, at 5:36 PM, Michael Bayer wrote:
Also, if you want to gain some understanding of objectstore, which
would be helpful if you have any ideas to offer, i just committed a
nicer "dependency dump" function. If you turn on
sqlalchemy.mapping.objectstore.LOG = True , youll see a "tas
s patch
as a reasonable starting point?
thanks,
daishi
---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as s
nse, since the cycle won't get
caught by the 't[0] is t[1]' check.
nevertheless, i also wonder if this is something
that you expect to support in the future?
sorry to be bringing up less common use-cases,
but sqlalchemy gives me the hope of leveraging
these features through python.
th
On Jan 20, 2006, at 6:12 AM, Jonathan Ellis wrote:
Incidently, the PEP has been updated following a python-dev thread in
which Ian Bicking noted that "I personally prefer cls, there are some
who use klass, and I haven't see class_ used." The PEP now recommends
the use of cls.
fwiw, as i unders
hi again,
i'm having some problems doing a .commit().
the test code is a very minor tweak on
the one from the previous backref thread,
and is included below.
the exception that i get is 'KeyError' -
i poked around via pdb, and see that there
are some odd things with the UOW dependencies,
but i g
On Jan 19, 2006, at 2:37 PM, [EMAIL PROTECTED] wrote:
illustrated in the sample code before, by
s/before/below/
---
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX sea
hi,
i'm trying to muddle my way through properties.py
in an attempt to understand what backref does, but
i'm getting a bit confused.
ultimately i'm just trying to clarify in my mind
what the differences are between using backref
and defining two separate properties. this is
illustrated in the sa
hi mike,
thanks as always for your thoughtful response.
(i hadn't expected a discussion on the little
side-comments in my example!)
On Jan 15, 2006, at 8:37 PM, Michael Bayer wrote:
ok...would we want to have all the information_schema tables available
in the ischema module ? if so, would we w
i believe the following is a fix
for a small buglet. (i encountered
a problem creating a table w/
foreign keys in a non-default schema)
d
Index: databases/postgres.py
===
--- databases/postgres.py (revision 817)
+++ databases/
On Jan 11, 2006, at 5:41 PM, [EMAIL PROTECTED] wrote:
i think at least using information_schema
as a baseline is probably useful; let me
go ahead and do what i was talking about
so you can at least have some code to look
at when thinking about how to generalize
to be cross-db.
i'm attaching bot
On Jan 10, 2006, at 7:49 PM, Michael Bayer wrote:
well I already have a few of the information_schema tables in a
package called sqlalchemy.databases.information_schema, and youre free
to embellish that package.
that's pretty much what i've been
doing so far ...
it might be more useful, c
On Jan 11, 2006, at 10:59 AM, David Geller wrote:
Trying to use help() under python interperter, and get error:
>>> import sqlalchemy
>>> help(sqlalchemy)
.
OSError: [Errno 20] Not a directory:
'/usr/local/lib/python2.4/site-packages/SQLAlchemy-0.91alpha-
py2.4.egg/sqlalchemy'
i believe
On Jan 10, 2006, at 6:38 PM, Michael Bayer wrote:
well if you are referring to the "information schema" that postgres
uses, only postgres has that specific schema represented within it
(and its also quite complicated). Mysql seems to support a slightly
different version of it and also one that
i just wanted to clarify since i realized that
the 'columns' in my example was rather ambiguous.
On Jan 10, 2006, at 5:19 PM, [EMAIL PROTECTED] wrote:
engine = create_engine('...')
c = engine.ischema.columns
c.select(c.c.table_name=='foo')
what i mean here is that 'c' is the *Table*
informatio
Hi,
I would like to make the information schema
more directly accessible. It appears that
currently this is only used for the
'reflecttable' method to introspect table
definitions.
Based on my current understanding of
sqlalchemy I think it would be reasonable
for SQLEngine instances to have an
'
hi again,
the docs indicate 'backreference' as the kwarg
to use to activate the feature as described in
"Useful Feature: Backreferences", but it looks
like the code expects 'backref'?
not sure if a doc fix or a code fix would be
preferred here.
thx,
d
, so I couldn't move
forward with debugging.
I am also using namespace schemas, and believe
that there is a problem with ForeignKey objects
not handling namespace schemas. I am attaching
a patch for that.
(I have simplified the attached test.py to not
use namespace schemas).
Thanks,
Dais
34 matches
Mail list logo
